In two experiments, two types of objective behavioral measures of impulsivity (rapid-decision vs. reward-directed) will be compared between three groups of adolescents with conduct disorder (CD). Specifically, three adolescent groups with CD will participate: those without histories of physical fighting (CDNonFight) and those with histories of either initiating fights non-impulsively (CDFightJPlan) or impulsively (CDFight/NonPlan). Preliminary studies indicate that behavioral measures of impulsivity differ not only between control and conduct disorder (CD) groups, but also CD patients with histories of physical fighting perform more impulsively than those without histories of fighting. This appears especially true for laboratory tasks that require rapid-decisional responses (as compared to tasks that require reward-directed responses). In Experiment 1, to compare behavioral measures of impulsivity, several task types will be used concurrently to determine which measures are most sensitive to group differences. In Experiment 2, performance feedback (reward, penalty, and combined reward/penalty) will be used to determine which task parameters improve group discrimination, and how each group's performance may be differentially modulated by feedback manipulations. In both experiments, behavioral impulsivity performance will be related to traditional clinical measures (parent, teacher, and self-report ratings) to provide further validation of these laboratory measures. The specific aims are: (1) to determine which laboratory task type (rapid-decision vs. reward-directed) best discriminates between controls and adolescents with CD. and between the CDNonFight, CDFight/Plan. and CDFight/NonPlan groups: (2) to determine how performance feedback differentially modulates impulsive responding and maximizes group discrimination: and (3) to provide further validation of laboratory measured impulsivity by relating performance to traditional clinical measures of impulsivity (prospective ratings from parents, teachers, and self-report). These studies will help meet our immediate goals of studying the basic mechanisms of impulsivity in these populations and to further develop measures of behavioral impulsivity. The development of objective repeatable (state-dependent) measures of impulsivity is needed to answer both basic and applied research questions. Our long-term goal is to determine how impulsivity relates to biological mechanisms and treatment prediction/outcome. Research focused on understanding the unique causal pathways that lead children to develop severe patterns of antisocial and aggressive behaviors will advance our treatment technology. Impulsivity is critical to many of these pathways, but unfortunately its conceptualization and operationalized in past research has been weak. This research will advance our understanding of the construct of impulsivity and will provide a firmer context within which future research can better study the role impulsivity plays in the development of severe conduct problems.